1.

Investment casting heat transfer (a) Using the grey body approximation with mold surface emissivity , the radiative heat ux from the mold surface to the surroundings is given by:
4 q = T2 ,

where T2 is the mold surface temperature. The radiative ux back from the (black, env = 1) environment to the mold surface is: 4 q = T3 , where T3 is the environment temperature. Since emissivity equals absorbtivity (in the grey body approximation), the amount of heat absorbed by the mold becomes:
4 q = T3 .

The net radiative heat transferred is thus:

4 4 qnet = (T2 T3 ).

(b) Let T1 be the temperature inside the mold (1450 K), T2 the outside mold temperature (unknown), and T3 the temperature of the large room (300 K), and with no information given, assume the large room is a black surface. The conductive ux through the mold is q=k T1 T2 , L

where L is the mold thickness given as 10 mm. If we add the convective ux to the net radiative ux from part 1a, the total heat ux away from the mold surface is:
4 4 q = (T2 T3 ) + h(T2 T3 ).

Set the convective and radiative plus convective uxes equal and solve for T2 : T1 T2 4 4 = (T2 T3 ) + h(T2 T3 ) L k kT1 4 4 T2 + + h T2 2 T3 + hT3 + =0 L L W W W 2.268 108 2 T 4 + (200 + 100) 2 T2 (183.7 + 90000 + 260000) 2 = 0 m K4 2 m K m Call the left side f (T ) and use Newtons method k Ti+1 = Ti f (Ti ) f (Ti )

Start with an initial guess halfway between T1 and T3 , and the iterations go (dont let the sub scripts confuse you, this is still T2 ): T0 = 800K, T1 = 1091K, T2 = 1068.863989K, T3 = 1068.673378K, T4 = 1068.673365K This is converging pretty quickly to T2 = 1068.7 K. (c) If the environment temperature is zero, then the convective heat ux is hT , so the total ux is: q = T 4 + hT Factoring out T gives q = (T 3 + h)T
So if we want a htotal for which q = htotal T , that would be T 3 + h.
1

(d) The Biot number is: Bi = Using the htotal above:

Bi = At 800K: Bi = At 1050K: Bi = At 1700K: Bi = hL k

(T 3 + h)L
k = 0.56

0.4 5.67 108 mW 4 (800K)3 + 100 mW 0.01m 2K 2 K

W 2 mK

0.4 5.67 108 mW 4 (1100K)3 + 100 mW 0.01m 2K 2 K

W 2 mK

= 0.66

0.4 5.67 108 mW 4 (1700K)3 + 100 mW 0.01m 2K 2 K

W 2 mK

= 1.05

All of these are in the mixed conduction/convection limited regime, which means temperature is somewhat nonuniform through the mold. Recall that for steadystate (heat) diusion, a Biot number of 1 means that the convection and conduction resistances are about the same, so the surface temperature is halfway between the mold interior temperature (the metal temperature) and the environment temperature; this is quite nonuniform for these purposes. (e) The temperature calculated in part 1b is close to the 1100K second case in part 1d. The Biot number is equal to the ratio of temperature dierences, which from part 1b gives: Bi = T1 T2 1300 1068.7 = = 0.30, T2 T3 1068.7 300

so the results are somewhat dierent. This is because the Biot number analysis underestimates the ux from the environment back to the mold.
However, qualitatively, this is still in the mixed convection/conduction limited regime, so the
temperature is somewhat nonuniform.
(f) We want to lower the Biot number to make this more uniform. We cant reduce L by making the mold thinner; if we try to make it stronger by making it thicker, then the temperature dierence across it will only increase and make failure more likely. We may be able to choose a dierent mold material with higher thermal conductivity. But the most eective thing to do is to try to reduce htotal , which is done in industry by wrapping the mold in a ceramic ber blanket, which reects much of the radiant heat back to the mold and creates a pocket of stagnant air to reduce the convective losses.